Cutting the cost of complexity

Across boardrooms and data centers, the systems that once promised speed and clarity are now clogging the arteries of business.

A new study from the IBM Institute for Business Value (IBV) says that automation is becoming the only way to tame the swelling intricacy of enterprise technology. Drawing on a survey of 680 IT leaders in 21 countries, the report concludes that companies using intelligent automation are cutting costs, improving security and driving growth. It portrays automation not as a futuristic add-on but as the discipline that keeps digital transformation from collapsing under its own weight.

“AI isn’t the problem anymore, it’s the solution to its own complexity,” said Jacob Dencik, IBM IBV’s Research Director and lead author of the study. “The challenge is that most organizations haven’t yet learned how to make it work for them.”

The report calls information technology “the powerhouse of the modern enterprise,” yet many IT departments remain burdened by old software, siloed data and uncoordinated purchasing. Average corporate IT spending has climbed 50% since 2023, from 6% to 9% of revenue. Two-thirds of those budgets now go toward transformation rather than upkeep. IBM’s researchers argue that the money delivers value only when organizations simplify architectures, consolidate oversight and embed automation throughout their technology stack.

Highly automated organizations in the study reported a 10% increase in revenue and a 28% reduction in IT costs tied to digital transformation efforts. They also saw a 16% faster time-to-market for new products and a 36% decline in downtime costs from cybersecurity incidents. These companies rely on AI-driven systems to monitor networks, fix code and allocate computing resources in real time. Data is automatically cleaned and standardized, and workflows run continuously without human intervention.

IBM defines intelligent IT automation as the use of artificial intelligence and machine learning to link processes across the enterprise so systems can act autonomously based on what they learn. When done correctly, automation creates a flywheel: better data and cloud strategies enable smarter automation, which in turn strengthens those same strategies. The most advanced organizations are already building toward self-directed operations in which AI agents manage infrastructure, compliance and security at scale.

Complexity remains the main obstacle. The report estimates that “shadow IT,” software and cloud services purchased and used outside official control, consumes about 24% of total IT spending, or roughly USD 192 million a year for a USD 10 billion firm. Technical debt from legacy systems adds further strain; three-quarters of executives expect it to reach high severity by 2026.

Automation helps reverse the trend. The survey shows that highly automated enterprises spend less overall while achieving better results, employing about 90 IT staff per billion dollars of revenue compared with 140 for less-automated peers. A key difference is cloud maturity. Firms that have completed at least 75% of their cloud migration are nine times more likely to fall into the “highly automated” group. Mature hybrid-cloud environments allow applications and data to move freely while maintaining visibility and security, preventing the duplication and waste that drive up costs.

IBM links these efficiencies to disciplined practices such as Infrastructure as Code, which standardizes deployments, and continuous testing that automatically tunes cloud configurations. The report notes that organizations adopting these methods translate savings directly into innovation budgets.

IBM says automation evolves in three stages. The first uses robotic process automation to handle repetitive work. The second adds prediction and pattern recognition through AI and machine learning. The third, now emerging, employs generative AI agents capable of complex reasoning and self-learning. Among the most advanced firms, 89% have implemented or are optimizing generative AI within their IT processes, compared with only 15% of laggards.

Data quality and integration determine whether those systems succeed. Half of all executives surveyed said disconnected technology limits how they can use data. Highly automated companies address this by using AIOps platforms that standardize management and eliminate redundancy. These systems give leaders continuous visibility into performance and allow them to prevent outages before they spread.

Security has improved in step. Two in three companies using AIOps reported smaller attack surfaces and better threat mitigation, while only 40% of less automated peers saw similar gains. Organizations that modernize their data infrastructure—by clearly defining which information is accessible to AI and which is not—are best positioned to realize the benefits of automation.

The report’s case studies illustrate the pattern. In Saudi Arabia, Al Rajhi Capital replaced fragmented systems with IBM middleware, unifying brokerage, asset management and investment banking services into a single app. Within a year, brokerage volume rose 40% and mutual-fund onboarding expanded tenfold. Inside IBM, the company’s own CIO organization used Apptio’s Technology Business Management model to analyze USD 2.5 billion in IT costs, retire redundant applications and reinvest in modernization. The result, the report says, was “cost and consumption transparency” and faster decision-making.

The study also finds that finance departments are becoming full partners in automation. In highly automated firms, finance teams are more likely to measure the impact of digital investments and apply those lessons to future budgets. This collaboration strengthens accountability and creates what IBM calls a “data-driven investment cycle,” in which savings from automation fund additional transformation. Organizations that optimize generative AI at scale report an average 90% return on digital transformation spending, compared with project-level gains for less mature peers.

IBM warns that without coordination, the very tools designed to simplify IT could recreate the same complexity they aim to eliminate. As each business unit experiments with its own AI applications, governance becomes critical. The report predicts that successful enterprises will consolidate oversight through centralized AI platforms that track which models and tools are used across departments, ensuring consistency and security.

Underlying the study is IBM’s “Hybrid by Design” framework, which connects technology modernization to business outcomes across three domains—product, integration and technology—spanning twelve capability areas from infrastructure to security. The framework promotes a nonlinear path to maturity in which progress in one domain accelerates the others. The authors of the study distill their findings into four priorities for leaders: modernize legacy applications and data, connect infrastructure systematically, integrate data and middleware, and infuse intelligence into every technology lifecycle.

Each step should be guided by clear business objectives and by the reinvestment of savings into new automation projects. The goal is not to replace human oversight but to channel it toward higher-value work.

The report’s conclusion is stark but hopeful. Technology will never be simple, but it can be made manageable. Intelligent automation, when applied deliberately, turns the sprawl of modern IT into a coherent system that can adapt to change.

“Automation works when people stop treating it as a project and start treating it as a mindset,” said Lisa Fisher, Research Leader at the IBM Institute for Business Value. “It’s about creating a foundation that can absorb change instead of fighting it.”